About Me
I am an Embedded Control Software Engineer at Caterpillar Inc., where I develop advanced control and predictive maintenance systems for mining site operations. Previously, I completed my Ph.D. in Electrical & Computer Engineering at Missouri University of Science and Technology (MST) under the guidance of Dr. Sarangapani.
My doctoral research focused on reinforcement learning-based optimal tracking control for nonlinear discrete-time systems, particularly for robotics and autonomous vehicles. A significant aspect of my work involved lifelong learning-based optimal controllers, enabling continuous improvement from past experiences. I also emphasized safety-aware and explainable AI to ensure the reliability and interpretability of autonomous decision-making systems.
At Caterpillar, I design and deploy MATLAB/Simulink Energy Management Systems (EMS) for mining operations, leveraging AI/ML techniques (including LSTM and CNN models) for load forecasting, fault detection, and image-based defect detection. My work integrates embedded software development, model-based design, and advanced analytics to enhance operational efficiency and predictive maintenance.
Beyond my core research, I explore machine learning applications in cyber-physical systems, including:
- Vision-based robotic manipulation and localization
- Motion planning & perception
- SLAM and mapping for real-world autonomous navigation
Research Interests
- Reinforcement Learning & Optimal Control: Developing safe and explainable deep reinforcement learning-based controllers for nonlinear, discrete-time systems, with real-world applications in robotics and autonomous systems.
- Navigation & Motion Planning: Designing adaptive and robust path optimization strategies for autonomous vehicles and mobile robots, enabling efficient navigation in off-road terrains (forests, deserts) and human-centered environments (sidewalks, crowded buildings).
- Perception & Sensor Fusion: Implementing multi-sensor fusion techniques (LiDAR, GPS, IMU) to enhance state estimation, localization, and tracking in dynamic environments.
- Artificial Intelligence in Autonomous Systems: Leveraging deep learning and AI-driven models to enhance decision-making and control in robotics, self-driving cars, and unmanned systems.
- Machine Learning for Control & Simulation: Integrating deep learning-based controllers with traditional model-based control (MPC, PID, fuzzy logic) for improved robustness in nonlinear and uncertain systems.
- Robotics & Autonomous Vehicles: Advancing motion control, planning, and reinforcement learning for humanoid robots, mobile manipulators, and self-driving platforms.
- Safety & Security in Nonlinear Systems: Developing safe reinforcement learning-based controllers with performance guarantees in critical autonomous operations.